An innovative approach combining industrial process data analytics and operator participation to implement lean energy programs: A Case Study

摘要

Energy costs for process-based industries amount to over 380 billion USD per year. In today’s economic climate, and considering environmental drivers, addressing energy efficiency is critical. Given that a typical plant captures and archives several thousand measurements per second, the challenge for industry today remains how to extract value from their “Big Data” to address energy efficiency. Continuous improvement programs aligned with lean manufacturing principles can optimize current assets leading to operational efficiency gains without capital investments. Combining advanced analytics and machine learning, with a strong involvement of plant staff and operators is key to deploying an Energy Management System (EnMS) compliant with ISO50001 that will help the plant to quickly achieve significant savings. This paper outlines the critical steps to implement an EnMS for a complex process: to diagnose energy consumption variability; identify energy consumption baseline; to engage all levels of production staff in root cause analysis workshops; and, to implement predictive models for real-time monitoring, decision support and performance reporting. The chemicals plant case study presented in this paper demonstrates how this approach can be applied to optimize steam production and distribution through improved operational management. This case realized operational savings of over 640 000 USD or 7000 tonnes of CO2 per year in gas consumption, representing a 15% reduction. Experience from this case emphasizes the importance of using plant monitoring to its full potential together with involvement of plant operators in order to understand key operational practices and to help promote an energy efficiency culture.